Interfaces between atmospheres and magnetospheres

Luke Moore, Boston UniversityMarina Galand, Imperial College London

Giant planet upper atmospheric physics, observations, and theory

Upper atmosphere “basics” Thermosphere, ionosphere, exosphere, homopause…

Generation of an ionosphere Photon absorption, particle precipitation

Ion production and loss

Remote ionospheric diagnostics Giant planet observations

Model-data comparisons Outstanding issues

Astrophysicists beware:

“H-two” = H2 ≠ HII“H-plus” = H+ = HII

Lick Observatory

Coma Surface-bound Exospheres

J. WilsonBaumgarder et al. (2008)

Dense Atmospheres

N2 atmospheres• Earth• Titan• Triton• Pluto

CO2 atmospheres• Venus• Mars• Pluto

H2/H/He atmospheres• Jupiter (P10/P11/V1/V2/Ulysses/Cassini/New

Horizons, Galileo)• Saturn (P11/V1/V2, Cassini)• Uranus (V2)• Neptune (V2)

And more…

• Enceladus• Io• Europa• Ganymede• Callisto

Lower atmosphere (meteorology)

Upper atmosphere (aeronomy)- Key transition region between

lower atmosphere and magnetosphere

- Energy and momentum sources:- EUV/FUV solar radiation- Energetic particles- Forcing from below (e.g.,

gravity waves)

I. Müller-Wodarg

Thermosphere:- Positive temperature

gradient- Collective (fluid) behavior

Exosphere:- Constant temperature

(“exospheric temperature”)- Infrequent collisions

kinetic particle behavior and escape

Reference altitude (0 km) = 1 bar level

Molecular diffusion

Convective mixing

Heterosphere

Homopause /Turbopause

Homosphere

Moses and Bass (2000)

Ionized part of upper atmosphere

Typically coincident with thermosphere, but

Present at any object with an atmosphere *

Ion densities << neutral densities Key layer for coupling between the upper

atmosphere and the magnetosphere

Closure of the magnetospheric current system

Conducting layer

Key source of heating of the high latitude upper atmosphere

Dominated by hydrogen: Distant: ~5.2, 9.5, 19, 30 and AU

(reduced solar insolation)

Fast rotators: ~9.925, 10.656, 17.24, and 16.11 hours/day

Widely varyingdipole alignments:

Jupiter Saturn Uranus Neptune

H2 89.8% 96.3% 82.5% 80.0%

He 10.2% 3.25% 15.2% 18.5%

CH4 1000 ppm 4500 ppm 2.3% 1.5%

Ionization thresholds:

H2: 15.43 eV (80 nm)

H: 13.60 eV (91 nm)

CH4: 12.55 eV (99 nm)

Solar EUV and X-ray (<10 nm) radiation:

Solar photon flux / (Sun-planet distance)2

Energetic particles from the space environment:

A few keV to a few 100s keV

13 eV ≈ 100 nm

True or False?

The higher the energy of a photon, the lower in altitude it will be absorbed.

The higher the energy of an electron, the lower in altitude it will be absorbed.

* Suprathermal electrons can be photoelectrons, auroral electrons, and/or secondary electrons

M. Galand

𝐼𝜆 = 𝐼𝜆𝑇𝑂𝑃𝑒𝑥𝑝

𝑛

𝜎𝜆𝑎𝑏𝑠(𝜆)

𝑧

∞

𝑛𝑛 𝑧′ sec 𝜒 𝑑𝑧′

Attenuated photon flux at wavelength l and at altitude z:

Optical depth t

solarzenithangle

Neutral number density

of constituent nSolar flux at top of atmosphere

Photoabsorption cross section

Ion (and photoelectron) production rate at wavelength l and at altitude z:

𝑃𝜆 𝑧 =

𝑛

𝜎𝑛𝑖𝑜𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛(𝜆)𝑛𝑛(𝑧)𝐼𝜆(𝑧) ∝ 𝐼𝜆

𝑇𝑂𝑃 ∝1

𝑑𝑠𝑢𝑛→𝑝𝑙𝑎𝑛𝑒𝑡2

1 Mb = 10-18 cm2

M. Galand

M. Galand

We’ve talked a lot about solar photons as sources of ionization. Why not stellar photons?

* Suprathermal electrons can be photoelectrons, auroral electrons, and/or secondary electrons

M. Galand

Thermal ion continuity equation

Photochemical equilibrium

When chemical processes dominate over transport (typically in lower ionosphere; e.g., terrestrial E region)

𝜕𝑛𝑖𝜕𝑡= 𝑃𝑖 − 𝐿𝑖 − 𝛻 ∙ 𝑛𝑖𝑢

Production

Loss

Transport

bulkvelocity

𝑃𝑖 = 𝐿𝑖

Radiative recombination (RR; atomic ions)

Charge exchange

Dissociative Recombination (DR; molecular ions)

𝑋+ + 𝑒− → 𝑋 + ℎ𝜈

𝐿𝑋+𝑅𝑅 = 𝛼𝑋+

𝑅𝑅𝑛𝑋+𝑛𝑒

𝑋+ + 𝑌 → 𝑋 + 𝑌+

𝐿𝑋+,𝑌𝐶𝐸 = 𝛼𝑋+,𝑌

𝐶𝐸 𝑛𝑋+𝑛𝑌

𝑋𝑌+ + 𝑒− → 𝑋 + 𝑌

𝐿𝑋𝑌+𝐷𝑅 = 𝛼𝑋𝑌+

𝐷𝑅 𝑛𝑋𝑌+𝑛𝑒 ≈ 𝛼𝑋𝑌+𝐷𝑅 𝑛𝑒

2

SLOW

FAST (typically)

FAST

If XY+ is dominant ion

Protonated molecular hydrogen

What is it?

H2+ accounts for ~90% of initial ion production H2+ + H2 H3+ + H R1

k1 = 2.0x10-9 cm3 s-1

H2+ rapidly converted to H3+

H3+ + e- neutrals R2

k2 ≈ 8.6x10-7 T-0.5 cm3 s-1

H+ becomes dominant due to slow RR loss and short day (rapid rotation)

H+ + e-H + hn R3

a3 ≈ 2x10-10 T-0.7 cm3 s-1

Initial theory therefore predicts: Predominantly H+ ionosphere with little

diurnal variation

H2+

H+

He+

CH3+

Moses and Bass (2000)

300+ additional reactions!

Simplified Schematic of HydrocarbonPhotochemistry atSaturn

Kim and Fox (1994)

Kim and Fox (2001)

Meteoroid ablation deposition leads to Mg/Mg+, Fe/Fe+, Si/Si+, O/O+, S/S+, C/C+, etc.

And many more…

Radio occultations Time delay and bending angle (a) provide

electron density vs. altitude

d

Withers et al (2014)

(1) Frequency residual vs. time

(2) Bending angle a vs. time (3) Refractivity vs. radius

(4) Electron density vs. radius

𝜇𝑒 − 1 = 𝜐𝑒 = −𝑛𝑒𝑒2

8𝜋2𝑚𝑒𝜖𝑜𝑓2

Saturn Electrostatic Discharges (SEDs) Broadband, short-lived, impulsive radio emission, ~10 hr periodicity

▪ Initially thought to originate in Saturn’s rings, later shown to be associated with powerful lightning storms in Saturn’s lower atmosphere

▪ Detected by Voyager and Cassini (~6 SED storms to-date, lasting weeks-months)

Observed low-frequency cutoff can be used to derive NMAX(t)

Powerful lightning also observed at Jupiter, but no “JEDs”

▪ Perhaps due to attenuation of radio waves by Jupiter’s ionosphere

H3+ observations Predicted to be a major ion in outer planet ionospheres

Plethora of H3+ emission lines available in IR, particularly through K-band (2-2.5 mm) and L-band (3-4 mm) atmospheric windows

To be continued in Part II…

6 Apr 1973 5 Sep 1977 20 Aug 1977

Pioneer 10 Pioneer 11 Voyager 1 Voyager 2 Galileo Cassini

15 Oct 199718 Oct 19893 Mar 1972

x 2 x 2

x 2

x 2

x 2

x 2

x 2

x 5*

x 59**

x 2

x 2

= 13

= 65

= 2

= 2

ingress (N) and egress (X) orbiters

* analyzed; ** taken to-date

Yelle and Miller (2004)

Galileo

Voyager 2

Voyager 1

Pioneer 10

NMAX ~ 105 cm-3

Fjeldbo et al (1975)

Eshleman et al (1979)

hMAX ~ 600-2000 km

= peak electron density

= altitude of NMAX

Lindal et al (1987)Lindal (1992)

Uranus Neptune

NMAX ~ 104 cm-3

hMAX ~ 1000-2000 km

NMAX ~ 103-104 cm-3

hMAX ~ 800-1500 km

VoyagerPioneer

NMAX ~ 104 cm-3 hMAX ~ 1000-2500 km

Kliore et al (1980)

Lindal et al (1985)

Nagy et al (2006)

DAWN

DUSK

Nagy et al (2006)

Galand et al (2009)

H+

H+ H+

H3+

Ionospheric model simulation

Cassini equatorial radio occultation

averages

H+

H3+

NeMoore et al (2004)

Ionospheric model simulation

NMAX ~ 103 cm-3

hMAX ~ 1200-2800 km

At Saturn’s equator:

Kliore et al (2009)

Moore et al (2010)

TEC = total electron content(1 TEC unit = 1016 cm-2)

Photoionization rates at Saturn peak near the equator and fall off with latitude. The observed electron density trend is exactly the opposite. What else might be happening?

Modeled NMAX larger than observed

Solution: convert long lived H+ into short lived molecular ions:

▪ Unconstrained charge exchange reaction

H+ + H2(n≥4) H2 + + H2 R4k4 ≈ 1x10-9 cm3 s-1 (Huestis, 2008)

▪ Water (or other external) influx

H+ + H2O H2O + + H R5k5 = 8.2x10-9 cm3 s-1

H2O+ + H2 H3O + + H R6k6 = 7.6x10-10 cm3 s-1

H3O+ + e- neutrals R7

a4 = 1.74x10-5 T-0.5 cm3 s-1

Modeled hMAX lower than observed

Above reactions act to slightly raise hMAX; in addition,

Forced vertical plasma drift?

Majeed and McConnell (1991)

A = Voyager radio occultationsB = nominal modelC = model fit with forced vertical drift +

enhanced H2(n≥4)

Majeed and McConnell (1991)

A = Voyager radio occultationsB = nominal modelC = model fit with forced vertical drift +

enhanced H2(n≥4) (left) or water influx (right)

Lyons (1995)Chandler and Waite (1986)

Uranus

Neptune

• No match to upper ionosphere• Produces low altitude layers using

meteoroid influx and vertical wind shears

• Exploration of effects of varying upper atmospheric temperatures, water and methane influxes, ionospheric outflows, and electron precipitations

LT of storm from images, angle of incidence a calculated from storm and Cassini position

)cos(

max,

a

pe

cutoff

ff

81/2

max,pee fN

Fischer et al (2011)

Moore et al (2012)

Majeed and McConnell (1996) Moore et al (2012)

Voyager result (A)

Various attempted

model fits (B-E)

Cassini results(dotted and dashed)

Various attempted model fits (grey)

Best model fit (solid lines)

Significant ionization enhancements required to match dawn-noon rise

Drastic losses required to match nighttime decline

Non-photochemical solution? Low altitude ion layers?

Ionization sources: EUV and X-ray solar photons, and

magnetospheric, energetic particles (dominant in auroral regions)

Giant planet ionospheres: Dominant ionization species (H2+) minor constituent after chemistry

Major ions:▪ H+: long-lived, minimal diurnal variation, subject to transport

▪ H3+: short-lived, strong diurnal variation, predominantly in photochemical equilibrium

▪ Hydrocarbon and metallic ions: extremely short-lived, bottomside “shoulder” of ionization

Unconstrained chemistry:▪ Populations of vibrational levels for H2 (in particular n≥4)

▪ Water (or other oxygen/metallic) influxes: variation with latitude, time, etc.

Remaining unknowns: Low altitude electron density layers: gravity waves or other vertical wind shear?

Origins of observed ionospheric structure and variability

Local time variations in ion and electron densities

SED explanation; lack of “JEDs”

Ionosphere-thermosphere-magnetosphere coupling at the giant planets

Auroral emissions Categories of aurora

UV vs. IR (i.e., H3+) aurora

Ionosphere-thermosphere-magnetosphere and solar wind coupling Saturn ring rain

The giant planet “energy crisis”▪ Upper atmospheric temperatures; heating sources

Ionospheric electrical conductivities

Future prospects Juno, JUICE, JWST, EChO, …

Aurora: photo-manifestation of the interaction between energetic

extra-atmospheric electrons, ions, and neutrals with an atmosphere

Unique and valuable remote diagnostic for the solar system

M. Galand

Saturn [HST]

Earth [ISS]

Color Ratio Earth Jupiter, Saturn

Two spectral bands chosen in:

One band strongly absorbed by:

Electron energy range covered:

Type of aurora identified:

N2 LBH

O2 (< 160 nm)

0.2 – 20 keV

Electron aurora (discrete only)

H2 Lyman and Werner

CH4 (< 140 nm)

~10 – 200 keV

Electron aurora (diffuse + discrete)

Identification of energetic particle type Assessment of Em and Qprec of energetic particles

Em = mean energy of precipitating particles (e.g., Maxwellian) Qprec = energy flux of precipitating particles

Similar techniques can be applied at various other planets with different limitations on the product (e.g., Fox et al, 2008).

Above tasks require comprehensive modeling support

(1) Emission from precipitating particles: radio and x-ray Radio emission generated by precipitating electrons as they are accelerated

into atmosphere along magnetic field lines▪ Originate in low density region above planet, near field-aligned potentials▪ Cause of auroral radio emission observed at all the giant planets (Zarka, 1998; Lamy et

al., 2009)

X-ray emission bremsstrahlung emission from high-energy precipitating particles scattered by the atmosphere (e.g., Jupiter)▪ Some electron driven bremsstrahlung present (e.g., Branduardi-Raymont et al., 2007),

but primarily due to highly charged heavy ions

Branduardi-Raymont et al (2007)

Jupiter [XMM-Newton EPIC]

Lamy et al (2009)

Saturn [Cassini RPWS]

(2) Atmospheric excitation Prompt emission resulting from atmospheric atoms and molecules

excited by precipitating particles The “classic” aurora (e.g., Earth)

▪ Similar on different planets, owing to composition differences

Brightest giant planet emissions: ▪ UV Lyman-a (121.6 nm); visible light Balmer series (e.g., 410.2 nm); UV H2 Lyman

and Werner bands (dominating over ~90-170 nm)

Provides instantaneous view of the particle precipitation process

Clarke et al (2009)

Jupiter Saturn

(3) Thermal auroral emission Produced from heating generated by atmosphere-magnetosphere

interaction Molecular hydrogen, hydrocarbons, and hydrogen ions emit IR

when thermalized to neutral atmosphere▪ Major heat sink in the upper atmosphere▪ H3+ most easily observed▪ Hydrocarbons provide majority of cooling

(4) Ionization aurora Ionization dominated by particle precipitation in auroral regions Due to long thermal timescales and short ionization timescales,

auroral structure is dominated by ionization, while overall brightness is dominated by temperature

Closely follows prompt UV auroral morphology; time and spatial lag due to H3+ recombination rates and temperature variations

First astronomical spectroscopic detection in the universe at Jupiter Auroral IR measurements with CFHT (Drossart et al., 1989)

Bright emission lines in K-band (2-2.5 mm) and L-band (3-4 mm) atmospheric windows Strong methane absorption in the L-band

▪ Therefore, at the giant planets (where H3 + is above the homopause), H3+ appears as bright emission above a dark background

Highly temperature dependent, T4

Can be used to derive ion temperatures, densities velocities

Important as a coolant, e.g.: Efficient thermostat at Jupiter Hot exoplanets with dissociated H2 lose a key cooling mechanism Connerney and Satoh (2000)

H3+ line-of-sight velocity and normalized intensity (NASA-IRTF)

Temperature Column Density

Lam et al (1997)

Wavelength (micron)

ring reflection methane

local extrema mirrored at magnetically conjugate latitudes, and also map to structures in the rings

First non-auroral detection of H3+

at Saturn

Keck observations: 2011

O’Donoghue et al (2013)

H2(n≥4) population

Family of solutions matching ring rain H3+

column density at -350

Global water influx at Saturn

Water influx mapped to ring plane

Moore et al (2014)

M. Galand

M. Galand

Heating sources: forcing from above and below Solar heating:

▪ excitation/dissociation/ionization and exothermic chemical reactions

Particle heating:▪ via collisions and chemistry

“Ionospheric joule heating”▪ via auroral electrical currents and ion-drag heating at high latitudes (e.g.,

Vasyliũnas and Song, 2005)

Dissipation of upward propagating waves▪ e.g., gravity waves, acoustic waves, etc. (Matcheva and Strobel, 1999; Hickey

et al., 2000; Barrow et al., 2012)

Earth (TW) Jupiter (TW) Saturn (TW)

Solar EUV/FUV heating* 0.5 0.8 0.2

Auroral particle/Joule heating* 0.08 100 5-10

* Strobel (2002)

Moore et al (2014)

All published temperatures at Saturn:IR = H3+ (IRTF/Keck); UV = solar and stellar occultations (Voyager/Cassini)

ring rain latitudes ring rain

latitudes

Mueller-Wodarg et al (2012)* Melin et al (2007)** Vervack and Moses (2013)

Auroral Joule heating sufficient to heat high latitude thermosphere

BUT polar sub-corotation due to auroral forcing (westward ion velocities) drives downward collapse and equator-to-pole circulation

Input of more magnetospheric energy only exacerbates the ion drag fridge effect (Smith et al, 2007; Smith and Aylward, 2009)

wind direction

Equinox simulation Tn

Yates et al (2014)

M. Galand

M. Galand

M. Galand

M. Galand

Conductance: mho (inverse of resistance, ohm backwards)Conductivity: mho/m 1 mho = 1 Siemens (the SI unit)

M. Galand

What might be causing the difference in conductance at Jupiter and Saturn?

M. Galand

Moore et al (2010)

Assume electron density is constant with altitude Assume ionosphere is composed of entirely one ion ~50% difference in derived Pedersen conductance, mostly due to

mass Pedersen layer near 1000 km at Saturn (~600 km at Jupiter)

Electron density profiles from Galileo, Voyager and Pioneer Background atm. and ion fractions based on model, scaled to observed Ne

0.001 mho 0.88 mho

Future Prospects

M. Galand

Analysis of auroral emissions: Valuable probe of ionosphere (IR), auroral particle source, ITM coupling, and magnetic field line

configuration

Jupiter: main oval driven by breakdown in co-rotation (Io)

Saturn: main oval mapped in the outer magnetosphere varying with solar wind conditions (Enceladus)

Uranus: solar wind dominated

Ionosphere-Thermosphere-Magnetosphere (ITM) coupling Ionospheric electrical conductances:

▪ Uncertainties in conductivities driven by limitation in electron (and ion) density estimates

▪ Differences in B field strength between Jupiter and Saturn yield significant conductance differences. Larger energy fluxes at Jupiter don’t compensate for the stronger B field. Implications for ITM coupling

Simulations:

▪ Critical to estimate the upper atmosphere response self-consistently

▪ Play a key role in efforts to understand underlying physics

Energy crisis remains unsolved:

▪ Investigate shorter timescales, E field variability unconstrained, role of waves, mid-latitude e-?

Lessons learned from Saturn useful for upcoming exploration of Jupiter (Juno/JUICE) and exoplanets(EChO, JWST)

M. Galand

M. Galand

M. Galand